Science-related attitudes are known to influence students’ science learning. However, the specific reasons of science learning differences based on science-related attitudes have not yet been identified. Therefore, this study aims to investigate the reasons for these differences in science learning at the conceptual level by analyzing the learned concept networks learned by students according to their levels of science-related attitude. This study compared and analyzed the learned concept networks according to students’ science-related attitude levels, alongside the inter-subtopic and intra-subtopic connection concept networks, across different attitude levels. A science-related attitude test was administered to 298 10th graders, and learning memory tests were conducted immediately and four weeks after the class to measure the level of science learning regarding topics related to ecosystem composition. The ecosystem composition topics were categorized into “the components of an ecosystem” and “the interactions between ecosystem components.” The results showed that students formed network of key concepts centered on similar, specific concepts regardless of topic, but the structure of these networks and the strength of connections between concepts varied. Students with lower-level group of science-related attitudes tend to remember fewer concepts, and subsequent learning significantly influences the conceptual structure of prior learning. Therefore, these findings suggest that for students in lower-level group, it’s necessary to present key concepts in a ‘chunk form’, linking them together, and consistently repeat key concepts. Furthermore, teaching and learning methods that clearly explain the connection between subsequent learning and prior learning are needed to enhance retention.